From: Giuseppe CAVALLARO <peppe.cavallaro@st.com>
To: Sripathi Kodi <sripathik@in.ibm.com>
Cc: ltp-list@lists.sourceforge.net, amrith <amrith@linux.ibm.com>,
Gowrishankar <gowrishankar.m@linux.vnet.ibm.com>,
Sripathi Kodi <sripathik@linux.vnet.ibm.com>
Subject: Re: [LTP] [PATCH] compiling realtime tests for SH.
Date: Fri, 13 Nov 2009 08:40:50 +0100 [thread overview]
Message-ID: <4AFD0D82.6050302@st.com> (raw)
In-Reply-To: <20091112141814.4599b83a@sripathi.in.ibm.com>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Sripathi Kodi wrote:
> On Mon, 09 Nov 2009 23:09:16 +0530
> Subrata Modak <subrata@linux.vnet.ibm.com> wrote:
>
>> Hi Giuseppe,
>>
>> Thanks for the patch. Meanwhile i did not find the RT guys saying
>> anything about this patch.
>>
>> Gowri/Amrith/Sripathi,
>>
>> Any views on this patch ?
>
> Hi Subrata,
>
> I think separating out the tests that need TSC from the others is
> the right thing to do. In future it may be useful to do something
> similar for atomic_* functions as well.
>
> In my opinion this patch would have been better split into 2
> parts: one that separates out TSC parts and the other that adds atomic
> functions for SH. That would make changelogs easier to search through
> in future.
if you like I can resend the patch splitted into two parts.
Let me know.
Regards,
Peppe
> Thanks,
> Sripathi.
>
>> Regards--
>> Subrata
>>
>> ** Also, this fails to apply:
>> patching file testcases/realtime/func/async_handler/async_handler_tsc.c
>> Hunk #1 succeeded at 46 (offset 1 line).
>> Hunk #2 succeeded at 193 (offset 2 lines).
>> patching file testcases/realtime/func/measurement/preempt_timing.c
>> patching file testcases/realtime/func/measurement/rdtsc-latency.c
>> patching file testcases/realtime/include/librttest.h
>> Hunk #2 succeeded at 98 (offset 1 line).
>> Hunk #4 succeeded at 138 (offset 1 line).
>> Hunk #5 succeeded at 337 (offset 8 lines).
>> patching file testcases/realtime/include/libtsc.h
>> ---
>>
>> On Wed, 2009-11-04 at 16:48 +0100, Giuseppe CAVALLARO wrote:
>>> Hoping somebody will find useful it.
>>>
>>> Regards
>>> Peppe
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> differences
>>> between files
>>> attachment
>>> (ltp-full-20090731-sh_RT.patch)
>>>
>>>
>>> [PATCH] compiling realtime tests for SH.
>>>
>>> This patch adds the atomi_add() define for __sh__;
>>> ltp realtime tests seem to show that it works.
>>>
>>> The patch also moves the tsc macros form the librttest.h
>>> to another header file (named libtsc.h).
>>>
>>> Without this I got the following error and no tests were built.
>>>
>>> [snip]
>>> In file included from librttest.c:43:
>>> ../include/librttest.h:127:2: error: #error
>>> ../include/librttest.h:169:2: error: #error
>>> In file included from librttest.c:43:
>>> ../include/librttest.h: In function ‘atomic_add’:
>>> [snip]
>>>
>>> The tsc macros header is only included in the following tests
>>> (thus where necessary):
>>>
>>> o async_handler_tsc.c
>>> o preempt_timing.c
>>> o rdtsc-latency.c
>>>
>>> Note1: instead of touching the make process the patch allows to build
>>> the tests above also for architecture that do not support tsc, yet.
>>> These tests will fail at run-time with ENOTSUP. A warning will
>>> appear while compiling as well.
>>> hmm, I do know if it is the right solution but it's simple and
>>> a good starting point for me.
>>>
>>> Note2: I've just started running this part of the ltp on stlinux
>>> targets so I cannot provide results and I don't know which
>>> tests will fail.
>>>
>>> Note3: realtime tests continues to build and apparently work on i386.
>>>
>>> Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
>>>
>>> diff -urN
>>> ltp-full-20090731/testcases/realtime.orig/func/async_handler/async_handler_tsc.c ltp-full-20090731/testcases/realtime/func/async_handler/async_handler_tsc.c
>>> ---
>>> ltp-full-20090731/testcases/realtime.orig/func/async_handler/async_handler_tsc.c 2009-11-04 15:55:58.000000000 +0100
>>> +++
>>> ltp-full-20090731/testcases/realtime/func/async_handler/async_handler_tsc.c 2009-11-04 15:39:21.000000000 +0100
>>> @@ -45,6 +45,7 @@
>>> #include <pthread.h>
>>> #include <librttest.h>
>>> #include <libstats.h>
>>> +#include <libtsc.h>
>>>
>>> #define HANDLER_PRIO 98
>>> #define SIGNAL_PRIO 99
>>> @@ -190,6 +191,11 @@
>>> int main(int argc, char *argv[])
>>> {
>>> int signal_id, handler_id;
>>> +
>>> +#ifdef TSC_UNSUPPORTED
>>> + printf("Error: test cannot be executed on an arch wihout
>>> TSC.\n");
>>> + return ENOTSUP;
>>> +#endif
>>> setup();
>>>
>>> rt_init("h", parse_args, argc, argv);
>>> diff -urN
>>> ltp-full-20090731/testcases/realtime.orig/func/measurement/preempt_timing.c ltp-full-20090731/testcases/realtime/func/measurement/preempt_timing.c
>>> ---
>>> ltp-full-20090731/testcases/realtime.orig/func/measurement/preempt_timing.c 2009-11-04 15:55:58.000000000 +0100
>>> +++
>>> ltp-full-20090731/testcases/realtime/func/measurement/preempt_timing.c 2009-11-04 15:39:26.000000000 +0100
>>> @@ -52,7 +52,7 @@
>>> #include <sys/mman.h>
>>> #include <stdint.h>
>>> #include <librttest.h>
>>> -
>>> +#include <libtsc.h>
>>>
>>> #define ITERATIONS 1000000ULL
>>> #define INTERVALS 10
>>> @@ -86,6 +86,12 @@
>>> struct sched_param param;
>>> cpu_set_t mask;
>>> int err;
>>> +
>>> +#ifdef TSC_UNSUPPORTED
>>> + printf("Error: test cannot be executed on an arch wihout
>>> TSC.\n");
>>> + return ENOTSUP;
>>> +#endif
>>> +
>>> max = avg = 0;
>>> min = -1;
>>> setup();
>>> diff -urN
>>> ltp-full-20090731/testcases/realtime.orig/func/measurement/rdtsc-latency.c ltp-full-20090731/testcases/realtime/func/measurement/rdtsc-latency.c
>>> ---
>>> ltp-full-20090731/testcases/realtime.orig/func/measurement/rdtsc-latency.c 2009-11-04 15:55:58.000000000 +0100
>>> +++
>>> ltp-full-20090731/testcases/realtime/func/measurement/rdtsc-latency.c 2009-11-04 15:38:49.000000000 +0100
>>> @@ -44,6 +44,7 @@
>>> #include <errno.h>
>>> #include <stdint.h>
>>> #include <librttest.h>
>>> +#include <libtsc.h>
>>>
>>> #define ITERATIONS 1000000
>>>
>>> @@ -101,6 +102,12 @@
>>> unsigned long long deltas[ITERATIONS];
>>> unsigned long long max, min, avg, tsc_a, tsc_b, tsc_period;
>>> struct sched_param param;
>>> +
>>> +#ifdef TSC_UNSUPPORTED
>>> + printf("Error: test cannot be executed on an arch wihout
>>> TSC.\n");
>>> + return ENOTSUP;
>>> +#endif
>>> +
>>> setup();
>>>
>>> rt_init("h",parse_args,argc,argv);
>>> diff -urN
>>> ltp-full-20090731/testcases/realtime.orig/include/librttest.h
>>> ltp-full-20090731/testcases/realtime/include/librttest.h
>>> ---
>>> ltp-full-20090731/testcases/realtime.orig/include/librttest.h 2009-11-04 15:55:58.000000000 +0100
>>> +++
>>> ltp-full-20090731/testcases/realtime/include/librttest.h 2009-11-04
>>> 14:21:22.000000000 +0100
>>> @@ -36,6 +36,7 @@
>>> * 2006-May-09: improved command line argument handling
>>> * 2007-Jul-12: Added latency tracing functions -- Josh Triplett
>>> * 2007-Jul-26: Renamed to librttest.h -- Josh Triplett
>>> + * 2009-Nov-4: TSC macros within another header -- Giuseppe
>>> Cavallaro
>>> *
>>>
>>> *****************************************************************************/
>>>
>>> @@ -96,36 +97,7 @@
>>> #define thread_quit(T) (((T)->flags) & THREAD_QUIT)
>>>
>>> #define PRINT_BUFFER_SIZE (1024*1024*4)
>>> -
>>> -/* TSC macros */
>>> #define ULL_MAX 18446744073709551615ULL // (1 << 64) - 1
>>> -#if defined(__i386__)
>>> -#define rdtscll(val) __asm__ __volatile__("rdtsc" : "=A" (val))
>>> -#elif defined(__x86_64__)
>>> -#define rdtscll(val) \
>>> - do { \
>>> - uint32_t low, high; \
>>> - __asm__ __volatile__ ("rdtsc" : "=a" (low),
>>> "=d" (high)); \
>>> - val = (uint64_t)high << 32 | low; \
>>> - } while(0)
>>> -#elif defined(__powerpc__)
>>> -#if defined(__powerpc64__) /* 64bit version */
>>> -#define rdtscll(val) \
>>> - do
>>> { \
>>> - __asm__ __volatile__ ("mfspr %0, 268" :
>>> "=r" (val)); \
>>> - } while(0)
>>> -#else /*__powerpc__ 32bit version */
>>> -#define
>>> rdtscll(val) \
>>> - do
>>> { \
>>> - uint32_t tbhi,
>>> tblo ; \
>>> - __asm__ __volatile__ ("mftbu %0" :
>>> "=r" (tbhi)); \
>>> - __asm__ __volatile__ ("mftbl %0" :
>>> "=r" (tblo)); \
>>> - val = 1000 * ((uint64_t) tbhi << 32) |
>>> tblo; \
>>> - } while(0)
>>> -#endif
>>> -#else
>>> -#error
>>> -#endif
>>>
>>> extern pthread_mutex_t _buffer_mutex;
>>> extern char * _print_buffer;
>>> @@ -135,7 +107,7 @@
>>>
>>> /* function prototypes */
>>>
>>> -/* atomic_add - add integer to atomic variable
>>> +/* atomic_add - add integer to atomic variable and returns a value.
>>> * i: integer value to add
>>> * v: pointer of type atomic_t
>>> */
>>> @@ -165,6 +137,20 @@
>>> : "cc", "memory");
>>>
>>> return t;
>>> +#elif defined(__sh__)
>>> + unsigned long t;
>>> +
>>> + __asm__ __volatile__ (
>>> +"1: movli.l @%2, %0 ! atomic_add_return \n"
>>> +" add %1, %0 \n"
>>> +" movco.l %0, @%2 \n"
>>> +" bf 1b \n"
>>> +" synco \n"
>>> + : "=&z" (t)
>>> + : "r" (i), "r" (&v->counter)
>>> + : "t");
>>> +
>>> + return t;
>>> #else
>>> #error
>>> #endif
>>> @@ -343,7 +329,6 @@
>>> */
>>> int ts_to_nsec(struct timespec *ts, nsec_t *ns);
>>>
>>> -/* return difference in microseconds */
>>> unsigned long long tsc_minus(unsigned long long tsc_start, unsigned
>>> long long tsc_end);
>>>
>>> /* rt_nanosleep: sleep for ns nanoseconds using clock_nanosleep
>>> diff -urN ltp-full-20090731/testcases/realtime.orig/include/libtsc.h
>>> ltp-full-20090731/testcases/realtime/include/libtsc.h
>>> ---
>>> ltp-full-20090731/testcases/realtime.orig/include/libtsc.h 1970-01-01
>>> 01:00:00.000000000 +0100
>>> +++
>>> ltp-full-20090731/testcases/realtime/include/libtsc.h 2009-11-04
>>> 15:41:03.000000000 +0100
>>> @@ -0,0 +1,72 @@
>>> +/******************************************************************************
>>> + *
>>> + * Copyright © International Business Machines Corp., 2006-2008
>>> + *
>>> + * This program is free software; you can redistribute it and/or
>>> modify
>>> + * it under the terms of the GNU General Public License as
>>> published by
>>> + * the Free Software Foundation; either version 2 of the License,
>>> or
>>> + * (at your option) any later version.
>>> + *
>>> + * This program is distributed in the hope that it will be useful,
>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
>>> + * the GNU General Public License for more details.
>>> + *
>>> + * You should have received a copy of the GNU General Public
>>> License
>>> + * along with this program; if not, write to the Free Software
>>> + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
>>> 02111-1307 USA
>>> + *
>>> + * NAME
>>> + * libtsc.h
>>> + *
>>> + * DESCRIPTION
>>> + *
>>> + * USAGE:
>>> + * To be included in some testcases.
>>> + *
>>> + * AUTHOR
>>> + * Darren Hart <dvhltc@us.ibm.com>
>>> + * Giuseppe Cavallaro <peppe.cavallarost.com>
>>> + *
>>> + * HISTORY
>>> + * It directly comes from the librttest.h (see its HISTORY).
>>> + *
>>> +
>>> *****************************************************************************/
>>> +
>>> +#undef TSC_UNSUPPORTED
>>> +
>>> +/* TSC macros */
>>> +#if defined(__i386__)
>>> +#define rdtscll(val) __asm__ __volatile__("rdtsc" : "=A" (val))
>>> +#elif defined(__x86_64__)
>>> +#define rdtscll(val) \
>>> + do { \
>>> + uint32_t low, high; \
>>> + __asm__ __volatile__ ("rdtsc" : "=a" (low),
>>> "=d" (high)); \
>>> + val = (uint64_t)high << 32 | low; \
>>> + } while (0)
>>> +#elif defined(__powerpc__)
>>> +#if defined(__powerpc64__) /* 64bit version */
>>> +#define rdtscll(val) \
>>> + do
>>> { \
>>> + __asm__ __volatile__ ("mfspr %0, 268" :
>>> "=r" (val)); \
>>> + } while (0)
>>> +#else /*__powerpc__ 32bit version */
>>> +#define
>>> rdtscll(val) \
>>> + do
>>> { \
>>> + uint32_t tbhi,
>>> tblo ; \
>>> + __asm__ __volatile__ ("mftbu %0" :
>>> "=r" (tbhi)); \
>>> + __asm__ __volatile__ ("mftbl %0" :
>>> "=r" (tblo)); \
>>> + val = 1000 * ((uint64_t) tbhi << 32) |
>>> tblo; \
>>> + } while (0)
>>> +#endif
>>> +#else
>>> +#warning TSC UNSUPPORTED
>>> +/* All tests will be compiled also for the
>>> + * architecture without TSC support (e.g. SH).
>>> + * At run-time these will fail with ENOTSUP.
>>> + */
>>> +#define rdtscll(val) do { } while (0)
>>> +#define TSC_UNSUPPORTED
>>> +#endif
>>> +
>>>
>>
>> ------------------------------------------------------------------------------
>> Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
>> trial. Simplify your report design, integration and deployment - and focus on
>> what you do best, core application coding. Discover what's new with
>> Crystal Reports now. http://p.sf.net/sfu/bobj-july
>> _______________________________________________
>> Ltp-list mailing list
>> Ltp-list@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/ltp-list
>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iEYEARECAAYFAkr9DXwACgkQ2Xo3j31MSSKO7wCgsH3KAqveTZfPlZhTeQDMLrMl
+w4An0XRbe7MZQa/Nk5Rs6D58rvIHIZk
=MRBR
-----END PGP SIGNATURE-----
------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-list
next prev parent reply other threads:[~2009-11-13 7:41 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-11-04 15:48 [LTP] [PATCH] compiling realtime tests for SH Giuseppe CAVALLARO
2009-11-09 17:39 ` Subrata Modak
2009-11-12 8:48 ` Sripathi Kodi
2009-11-13 7:40 ` Giuseppe CAVALLARO [this message]
2009-11-16 8:13 ` Subrata Modak
2009-11-16 10:53 ` Giuseppe CAVALLARO
2009-11-16 13:39 ` Subrata Modak
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4AFD0D82.6050302@st.com \
--to=peppe.cavallaro@st.com \
--cc=amrith@linux.ibm.com \
--cc=gowrishankar.m@linux.vnet.ibm.com \
--cc=ltp-list@lists.sourceforge.net \
--cc=sripathik@in.ibm.com \
--cc=sripathik@linux.vnet.ibm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox